Interactions of Different Species of Phytophthora with Cacao Induce Genetic, Biochemical, and Morphological Plant Alterations

Diseases associated with Phytophthora cause considerable losses in cocoa production worldwide. Analyzing genes, proteins, and metabolites involved in Theobroma cacao’s interaction with Phytophthora species is essential to explaining the molecular aspects of plant defense. Through a systematic literature review, this study aims to identify reports of genes, proteins, metabolites, morphological characteristics, and molecular and physiological processes of T. cacao involved in its interaction with species of Phytophthora. After the searches, 35 papers were selected for the data extraction stage, according to pre-established inclusion and exclusion criteria. In these studies, 657 genes and 32 metabolites, among other elements (molecules and molecular processes), were found to be involved in the interaction. The integration of this information resulted in the following conclusions: the expression patterns of pattern recognition receptors (PRRs) and a possible gene-to-gene interaction participate in cocoa resistance to Phytophthora spp.; the expression pattern of genes that encode pathogenesis-related (PRs) proteins is different between resistant and susceptible genotypes; phenolic compounds play an important role in preformed defenses; and proline accumulation may be involved in cell wall integrity. Only one proteomics study of T. cacao-Phytophthora spp. was found, and some genes proposed via QTL analysis were confirmed in transcriptomic studies.


Introduction
Diseases caused by species of oomycetes of the genus Phytophthora are responsible for substantial losses in the worldwide production of cocoa (Theobroma cacao L.) [1]. Among the most prevalent species, P. palmivora (E. J. Butler) affects cocoa plantations globally, while P. megakarya (Brasier and Griffin), which occurs only in Africa, affects the production of this cultivar. Other species such as P. megakarya [2], P. capsici (Leonian, 1922), and P. citrophthora (R.E. Sm. & E.H. Sm. Leonian) cause significant losses in cocoa production in Central and South America [1]. Furthermore, a new species, P. theobromicola, was recently discovered in cocoa beans collected from different farms in Bahia, Brazil [3].
The typical symptom caused by Phytophthora spp. is the rotting of the fruit's skin. This rot starts with small, hard, dark lesions that overgrow in a few days, covering the entire fruit surface and internal tissues, including the kernels [4]. Other parts of the plant may also be affected, although less frequently, such as the stem, flower cushions, leaves, and roots [5].
The control of diseases caused by Phytophthora spp. mainly involves cultural techniques for disinfecting tools and applying chemicals. However, due to factors such as the

String Construction and Selection for Database Searches
The objective of the systematic literature review was used as a basis to construct the main question and conduct the searches. The main question of this study is: "Which genes, proteins, metabolites, morphological characteristics, and molecular and physiological processes of T. cacao are involved in its interaction with Phytophthora sp.?" From this central question, keywords were selected to construct the strings (Table 1), which were tested in the database for academic research Scopus (https://www.scopus. com, accessed on 1 January 2023). The string that would be used in other databases for academic research was selected according to the following criteria: (a) the number of articles considered useful for answering the main question according to the abstract of the first 20 articles that appeared in the list; (b) whether or not sentinel articles were included in the first 20 articles that appeared in the list; and (c) the specificity of the string (Table 1).

Literature Search
The string selected and used in the search was {TITLE-ABS-KEY[(cacao OR cocoa) AND (resistance OR responses OR interaction OR defense OR protein* OR gene*) AND (Phytophthora)]}. In some cases, this string was adapted according to the rules of each site for literature searches (full articles, reviews, short communications, scientific notes). The search was carried out in the academic databases Scopus (https://www.scopus. com, accessed on 1 January 2023), Web of Science (https://www.webofknowledge.com, accessed on 1 January 2023), Science Direct (https://www.sciencedirect.com, accessed on 1 January 2023), PubMed (https://pubmed.ncbi.nlm.nih.gov, accessed on 1 January 2023), and Scielo (https://scielo.org, accessed on 1 January 2023). The string keywords were translated into Spanish and Portuguese for the latter database. Only articles from journals indexed in scientific databases were selected.  [24] and Changes in Gene Expression in Leaves of Cacao Genotypes Resistant and Susceptible to Phytophthora palmivora Infection [25] present in the first 20 items in the list). The string highlighted in bold was selected for the literature search. * = Boolean operator: the search engine will return any word that begins with the root/stem of the word truncated by the asterisk.

Literature Selection
Initially, a protocol was filled out with the key questions that guided the research (Supplementary S1). The identification, selection, and data extraction stages were performed using the tool StArt (State of the Art through systematic review) [26] v.2.3.4.2. For the selection stage, inclusion and exclusion criteria were established, and the option to automatically identify duplicate articles was used. Inclusion criteria: studies addressing the molecular, biochemical, physiological, and/or morphological responses of cocoa to any species of Phytophthora; studies reporting genes, transcripts, proteins, and/or metabolites identified in cocoa in response to interaction with any species of Phytophthora; studies addressing the characterization of T. cacao genes or proteins, but in transgenic plants (whether or not T. cacao) that demonstrate response or resistance to Phytophthora sp. Exclusion criteria: studies reporting populations, accessions, and/or genotypes with resistance to brown rot, but ignore the analysis methodologies that explain resistance; studies addressing only Phytophthora sp., but relate resistance to another phytopathosystem; studies addressing only the genetics, proteomics, and/or physiology of Phytophthora sp. without relating them to cocoa; studies reporting only QTLs or molecular markers in cocoa related to resistance to Phytophthora sp., but without reporting specific genes; studies reporting methodologies for screening resistance of T. cacao to Phytophthora sp; studies addressing the interaction of T. cacao with any species of Phytophthora; studies addressing the influence of external treatments (biological or chemical) on the interaction between T. cacao and Phytophthora sp.

Data Extraction
Once the articles were selected according to the inclusion and exclusion criteria, the data extraction stage was performed. For this purpose, blanks to fill in were established, corresponding to the general information about each study, as shown in Table 2. For the blanks "type of analysis" and "plant tissue studied", the options were fixed, and more than one option could be selected. In the blank "species of Phytophthora used in the study" and for each of the "reported elements", a space for free text was provided. In addition, blanks of elements reported for both resistant and susceptible genotypes were made available. The year of publication and the country where the study was conducted were also extracted.

Subsection Literature Search and Selection
A total of 490 articles were returned using the refined string in the five academic databases. About 80% of these were returned by Web of Science and Scopus, and the rest by the other three chosen databases ( Figure 1A).
Of the total number of articles found, more than half were rejected to move on to the information extraction stage, according to the exclusion criteria ( Figure 1B,C). Most were rejected because they do not address the interaction of T. cacao with some species of Phytophthora and only report populations/accessions/genotypes with resistance to brown rot, disregarding molecular analyses. Moreover, they only addressed the genetics, proteomics, or physiology of Phytophthora spp. without relating them to cocoa ( Figure 1C).
In turn, more than 40% of the studies were detected as duplicates, and only 35 were accepted for the data extraction stage, according to the inclusion criteria ( Figure 1B). Most accepted articles address the molecular, biochemical, physiological, and/or morphological responses of T. cacao to any Phytophthora spp.

Countries of Origin and Year of Publication of Papers Accepted for Data Extraction
Most papers accepted were developed in African countries and the USA (Figure 2). In South America, the country with the most articles published within the inclusion criteria was Trinidad and Tobago, followed by Brazil. More than 60% of the articles were published between 2011 and 2022 ( Figure 2); the oldest was published in 1961.

Data Extraction
In the information extraction stage, blanks were established to be completed with the general attributes for each accepted article ( Figure 3). Regarding the species of Phytophthora, the most widely selected to study the interaction with T. cacao were P. megakarya and P. palmivora. In turn, the most examined tissue was the leaf, followed by the fruit. In most accepted studies, the reported types of T. cacao elements involved in the interaction with Phytophthora sp. were genes and metabolites. Finally, regarding the type of analyses, the majority were on biochemistry, followed by genetics and transcriptomics. Only one study on proteomics [24] was within the inclusion criteria for articles accepted for data extraction. Of the total number of articles found, more than half were rejected to move on to the information extraction stage, according to the exclusion criteria ( Figure 1B,C). Most were rejected because they do not address the interaction of T. cacao with some species of Phytophthora and only report populations/accessions/genotypes with resistance to brown rot, disregarding molecular analyses. Moreover, they only addressed the genetics, proteomics, or physiology of Phytophthora spp. without relating them to cocoa ( Figure 1C).
In turn, more than 40% of the studies were detected as duplicates, and only 35 were accepted for the data extraction stage, according to the inclusion criteria ( Figure 1B). Most accepted articles address the molecular, biochemical, physiological, and/or morphological responses of T. cacao to any Phytophthora spp.

Elements That Participate in the Interaction with Species of Phytophthora
Data were extracted from the twelve studies that reported genes involved in the interaction between cacao and Phytophthora sp. to create a table (Table S1) with the 657 genes and 15 groups of transcripts of T. cacao considered the most relevant, most of which were related to defense.
Only one proteomics study on the interaction between cacao and Phytophthora sp. met the inclusion criteria [24]. This study identified 37 proteins in the resistant genotype and 39 proteins in the susceptible genotype through 2D-SDS-PAGE and mass spectrometry ( Table S2). Several of these proteins were identified more than once in different spots.
The other elements of T. cacao reported in the interaction with Phytophthora spp. are shown in Table 3. The most frequently reported enzyme was polyphenol oxidase. Almost all the secondary metabolites mentioned were phenolic compounds and their derivatives. In addition, four transiently expressed phosphatidylinositol-3-phosphate binding proteins were reported.

Data Extraction
In the information extraction stage, blanks were established to be completed with the general attributes for each accepted article ( Figure 3). Regarding the species of Phytophthora, the most widely selected to study the interaction with T. cacao were P. megakarya and P. palmivora. In turn, the most examined tissue was the leaf, followed by the fruit. In most accepted studies, the reported types of T. cacao elements involved in the interaction with Phytophthora sp. were genes and metabolites. Finally, regarding the type of analyses, the majority were on biochemistry, followed by genetics and transcriptomics. Only one study on proteomics [24] was within the inclusion criteria for articles accepted for data extraction. The identified molecular responses include signal transduction, response to various stresses and stimuli, and biosynthesis of phenylpropanoids, ethylene, and jasmonic acid. The reported morphological characteristics were almost always from the fruit of T. cacao, while only one study reported the morphological characteristics of the stem.
The only physiological response studied in the selected papers was moisture content. Other reported elements that possibly participate in the interaction between T. cacao and Phytophthora sp. were amino acids, lignin, carbohydrates, wax, caffeine, and synthetic peptides.

Elements That Participate in the Interaction with Species of Phytophthora
Data were extracted from the twelve studies that reported genes involved in the interaction between cacao and Phytophthora sp. to create a table (Table S1) with the 657 genes and 15 groups of transcripts of T. cacao considered the most relevant, most of which were related to defense.
Only one proteomics study on the interaction between cacao and Phytophthora sp. met the inclusion criteria [24]. This study identified 37 proteins in the resistant genotype and 39 proteins in the susceptible genotype through 2D-SDS-PAGE and mass

Few Primary Studies Address the Molecular and Morphological Aspects of the Interaction of T. cacao with Phytophthora spp.
Diseases associated with species of Phytophthora cause greater losses in cocoa production than any other disease [1]. Therefore, it is common for these species to be highlighted in abstracts of studies on cocoa in different areas, including its interaction with other pathogens. In turn, some studies address the morphological and molecular characteristics of species of Phytophthora and highlight the most important hosts, such as cacao. Thus, many articles were rejected by the exclusion criteria because they needed to address the interaction between T. cacao and Phytophthora ( Figure 1C), although these keywords were included in the title or abstract.
The low number of articles accepted according to the inclusion criteria (35 out of 490 found via the keywords; Figure 1D) indicates that there are few primary studies analyzing the molecular and morphological aspect of the interaction of T. cacao with Phytophthora spp. Additionally, this reflects the availability of analysis technologies over time, as the number of articles that met the inclusion criteria doubled from 2002 to 2020 ( Figure 2). During these years, technologies for molecular analysis, such as sequencing techniques, evolved rapidly and became more accessible over time. Moreover, the cocoa genome was sequenced and published only in 2010 [10], which allowed the advancement of research in the area of genetics and transcriptomics with this organism.
The two species most widely addressed in the study of interaction with cocoa in the articles accepted for data extraction were P. palmivora and P. megakarya (Figure 3). These species were probably chosen because P. palmivora causes huge losses worldwide, and P. megakarya affects the main cocoa-producing countries in Africa, the continent with the highest cocoa production in the world [1]. This also explains why African countries have developed more research on the biochemical, genetic, transcriptomic, and morphological interaction between cocoa and Phytophthora spp., together with the USA (Figure 2), which is a global leader in science and technology [45].
Regarding the type of analysis, most of the accepted articles were in biochemistry ( Figure 3). This may be due to the accessibility and speed of the tests required for biochemical analysis. Moreover, it provides relevant information on metabolic pathway products' role in plant defense. However, despite having more biochemical studies among the accepted works, there were only 32 records of metabolites (mainly phenolic compounds) and only 5 of enzyme activities. On the other hand, more than 600 records of genes were involved in the interaction of T. cacao with Phytophthora spp. (Tables 3 and S1), which highlights the importance of complex gene regulation processes in the plant's defense response.
Although the fruit is part of the plant most affected by Phytophthora spp. [4], there were more records of the use of leaves in the studies accepted for data extraction (Figure 3). Leaf bioassays have allowed the early and rapid selection of cocoa cultivars resistant to brown rot [18]. Therefore, leaf assays are a faster and more efficient way to study the effects of the pathogen on the cacao tree.

Expression Patterns of Pattern Recognition Receptors (PRRs) Are Crucial in Cacao Resistance to Phytophthora spp.
In all, 81 different T. cacao genes encoding probable disease-resistance proteins have been reported (Table S1) [21,36,37,46]. These genes were classified as possible pattern recognition receptors (PRRs) [21]. Among these, five genes had their expression repressed 24 h after inoculation with P. megakarya, in genotypes showing resistance, and one had its expression induced in the susceptible genotype 72 h after inoculation.
The gene expression before inoculation in resistant and susceptible genotypes of T. cacao was compared to identify genes involved in basal defense [21]. As a result, 20 genes encoding different probable defense-associated receptors were expressed in greater amounts in the resistant genotype than in the susceptible genotype. In the susceptible genotype, however, 27 genes that encode different probable receptors associated with defense were expressed in greater quantity when compared to the resistant genotype. Interestingly, only the susceptible genotype exhibited the leaf rust 10 disease-resistance loci receptor-like protein kinase-like sort, which is involved in abscisic acid (ABA) signaling [47], and only the resistant genotype exhibited inactive receptors (Table S1).
PRRs are known to be the first step of the plant defense response, as they recognize molecular patterns of the pathogen or cell wall damage [48]. Different expression patterns of PRRs were observed between the resistant and susceptible genotypes [21]. These differences emphasize the importance of the differentiated set of receptors for pathogen recognition in the plant, as the pathogen can secrete effectors that block or prevent recognition by certain PRRs [48].

Expression Patterns of Genes Encoding Pathogenesis-Related (PRs) Proteins Differ between Resistant and Susceptible Genotypes
In all, 109 genes encoding pathogenesis-related (PRs) proteins were reported. These show different expression patterns according to characteristics such as (i) the degree of resistance of the T. cacao variety, (ii) the time after infection, and (iii) the studied species of Phytophthora (Table S1) [8,21,23,36,37,49,50].
PR proteins are also called defense-related inducible proteins [51]. Signaling molecules induce PR in the cell, such as salicylic acid, jasmonic acid, and ethylene, after the pathogen has entered the cell. In a cacao-pathogen interaction research [49], microarrays were used to evaluate the gene expression level of the 17 families of PRs after inoculation with P. palmivora in a variety of T. cacao considered resistant to brown rot by [41]. In the study, 76 genes encoding PRs were induced. The family with the highest number of induced genes was PR-9, with twelve peroxidases, followed by PR-10 (eight ribonucleases), and PR-3 (eight chitinases).
However, the ability of PRs to prevent or reduce disease in plants depends on several factors, such as plant and pathogen species and the type of protein [51]. Inducing a large number of PRs does not imply that the plant will develop resistance to the pathogen. For example, the brown rot susceptible genotype had more PR family members induced after treatment with salicylic acid [52]. Interestingly, the biosynthesis and action routes of ethylene and jasmonic acid were induced in the Phytophthora spp. [37]. As discussed above, the compounds induce PR expression.
Expression of PRs also varies according to tissue and time of infection, where the expression of genes encoding a PR-1 changed depending on the tissue, with increased expression in the leaf and decreased expression in the fruit at 24 HAI [37].

Phenolic Compounds Play an Essential Role in Preformed Defenses
Baseline differences were observed in the gene expression of resistant and susceptible genotypes of cocoa inoculated with P. megakarya [21]. More dramatically expressed genes in the resistant genotype include Anthocyanidin 3-O-glucosyltransferase 5, which is involved in the biosynthesis of anthocyanin-containing compounds. Anthocyanins are phenolic compounds that may play an important role in plant defense against various pathogens [53]. Another gene with a higher basal expression in the resistant genotype encodes a detoxification protein [42]. This protein is also involved in the biosynthesis of phenolic compounds (flavonoids and proanthocyanidins) [54].
In this regard, phenolic compounds were the most reported metabolites in accepted papers that studied the T. cacao and Phytophthora spp. (Table 3). The profiles of basal metabolites in cocoa genotypes resistant and susceptible to Phytophthora species show essential differences. The resistant genotype accumulated higher levels of clovamide and several other amides of hydroxycinnamic acid (HCAAs) compared to the susceptible genotype, with clovamide being the most abundant metabolite [22].
This information suggests that the basal content of phenolic compounds and the capacity to synthesize these compounds in the initial stage of the infection are critical to the resistance of T. cacao to Phytophthora spp.

Possible
Gene-to-Gene Interaction May Be a Factor in the Resistance of T. cacao to Phytophthora spp.
According to previous studies, no gene-to-gene interactions have been demonstrated between cocoa and any of its pathogens [21]. However, a gene encoding a probable diseaseresistance protein containing the NB-ARC domain was identified with much higher basal expression in the resistant genotype compared to the susceptible genotype ( [21]; Table S1). In potatoes, this type of protein provides resistance to Phytophthora infestans carrying the Avr1 avirulence gene [55,56]. Therefore, this gene containing NB-ARC may be fundamental in the incompatible interaction of the resistant cacao genotype with some pathotypes of Phytophthora spp.

Genes Proposed via QTL Analysis Were Confirmed in Transcriptomic Studies
Some genes close to the QTL regions in cacao associated with resistance to P. palmivora, P. citrophthora, and P. capsici were discovered using chromosomal locations of the markers as a reference [8]. Interestingly, several genes proposed in this study were identified through transcriptomic analyses in other studies (Table S1), as were genes that encode glutaredoxin-C11 [21], calcium-binding protein CML23 [21], an F-box protein [21], an F-box/LRR-repeat protein [21], an LRR receptor-like serine/threonine-protein [21,37], a peroxidase [21,49], a serine-threonine protein kinase [21], and a zinc finger CCCH domain-containing protein [21]. However, the localized genes shown previously in the QTL regions do not contain the genome locus ID. When looking for the indicated location on each chromosome, the genes found in the cacao genome bank do not coincide with the gene name in the table, possibly due to a recent update to these loci.
A similar situation was identified in the other study [46], where three of the 324 candidate genes for recognition and activation of responses to the pathogen, identified in 20 QTL regions associated with resistance to P. palmivora, coincide with genes identified as differentially expressed [21], namely Tc10v2_g012100 (putative cc-nbs-lrr resistance protein), Tc10v2_g012750, and Tc10v2_g012770 (putative receptor-like protein 12). The two genes corresponding to receptor-like protein 12 were reported to be up-regulated in the resistant genotype compared to the susceptible genotype before inoculation with P. megakarya. In turn, the gene corresponding to the putative CC-NBS-LRR resistance protein was downregulated in the resistant genotype 24 h after inoculation with P. megakarya [21].

Accumulation of Proline May Be Involved in Cell Wall Integrity
Proline is an amino acid that accumulates under stress and normal conditions as a beneficial solute in plants and plays an essential role in cell wall proteins [57]. Proline stood out as a differential element between resistant and susceptible genotypes in two studies ( [34,42]; Table 3).
Furthermore, a gene encoding an ornithine decarboxylase, involved in the biosynthesis of a proline precursor [58], was reported to be induced at 48 HAI with P. megakarya ( [59]; Table S1). Regarding its importance in cell wall proteins such as arabinogalactan-proteins (AGPs), a gene encoding a Fasciclin-like arabinogalactan protein 2, rich in proline, was reported to be induced, 72 HAI, in a genotype resistant to P. megakarya ( [21]; Table S1). This information suggests that the induction of proline accumulation in resistant genotypes may be involved in the rapid recovery of the cell wall, which would be a factor in cocoa resistance to Phytophthora spp. and is consistent with the high lignin content found in resistant genotypes ( [7,39]; Table S1).

A Study at the Level of Proteomics Reveals Crucial Proteins in the Interaction of Cacao and P. palmivora
Despite the great economic losses due to diseases caused by Phytophthora spp. in T. cacao, only one study involving the proteomic analysis of this phytopathosystem was found (Figure 3; [24]). In the study, the proteomic profile of leaves of the genotypes of T. cacao PA150 (resistant) and SIC23 (susceptible) 48 HAI with the pathogen P. palmivora was compared with the respective controls inoculated with sterile distilled water [24]. The lipoxygenase proteins, 2-methylene-furan-3-one, co-chaperonin CPN20, and a probable CC-NBS-LRR stand out for having differentiated accumulation in the inoculated resistance genotype and are probably involved in the resistance of cocoa to P. palmivora. Interestingly, a gene encoding a CC-NBS-LRR was identified as expressed basally in the susceptible genotype [21]. In our analysis, this protein had a different accumulation pattern between the resistant and susceptible genotypes.
Among the transcriptomic studies reported in the present study, only one confirmed the expression of proteins through the Western blot technique [35]. Among these proteins, the four phosphatidylinositol-3-phosphate binding proteins, transiently expressed in T. cacao leaves, reduced infection by P. tropicalis and P. palmivora due to their ability to block the effector's entry of the pathogen to the cell. Despite the abundance of data about genes differentially expressed in cocoa before and during its interaction with Phytophthora spp., studies at the proteomic level are especially relevant since gene expression is regulated at different stages through processes that occur after the synthesis of transcripts such as post-transcriptional, translational, post-translational regulation, and protein degradation [60,61].

Conclusions
Expression patterns of pattern recognition receptors (PRRs) and a possible gene-forgene interaction participate in cacao resistance to Phytophthora spp.
The expression pattern of genes encoding pathogenesis-related proteins (PRs) is different between resistant and susceptible genotypes.
Phenolic compounds play an important role in preformed defenses. Proline accumulation may be involved in cell wall integrity. Genes proposed via QTL analysis encoding a glutaredoxin-C11, a calcium-binding protein CML23, an F-box protein, an F-box /LRR-repeat protein, an LRR receptor-like serine/threonine protein, a peroxidase, a serine-threonine protein kinase, and a zinc finger CCCH domain-containing protein were confirmed in transcriptomic studies.